A prior post shows that a much smaller estimate of WIFIA’s likely interest rate re-estimate losses can be developed with a plausible-sounding methodology.  The realpolitik purpose is to provide the basis for a defensive narrative in case of ideological or political attack.

However, there might be another line of attack on WIFIA’s cost that doesn’t rely on the sticker shock of large numbers.  This one would be based on concepts that are close in spirit to the justifications put forward in 2020 by those in Congress who wanted to defund the WIFIA program about delayed criteria for lending to federally involved projects.  There was no discussion about the economic impact of the criteria (there isn’t any), the number of WIFIA loan applications that would be affected (very few – one or two annually) or the cause of the delay (almost certainly unintentional, perhaps pandemic-related).  Instead, the narrative centered on how WIFIA was violating sacred federal budgeting principles, opening the door to all sorts of abuse, disregarding Congressional intent, etc. etc.  None of it was particularly true or important, but that didn’t matter – the attack nearly succeeded.

A new version of the 2020 tactic would point out (with appropriate rhetoric) that WIFIA’s FYE 2022 portfolio will indisputably cost more than the discretionary appropriations allocated for it, and that oversight agencies should investigate the Program’s compliance with the Anti-deficiency Act.

The Anti-deficiency Act’s name is scary enough for plenty of soundbites, and the basic assertion is in fact correct, regardless of how the scale of the cost numbers might be packaged.  But on the surface, WIFIA’s defense is succinct and bulletproof:  FCRA law requires that the cost of interest rate estimates go into an off-budget account (for which permanent indefinite budget authority is provided) and the Anti-Deficiency Act explicitly excludes shortfalls of discretionary appropriations authorized by law.  Case closed, no?

It would seem so, and an attack along these lines is perhaps unlikely.  But I’m not entirely certain that the Program is permanently exempt from questions about the FYE 2022 portfolio’s obvious deficiency of discretionary appropriations.  There’s no doubt that the letter of the relevant law is being followed scrupulously, but is that true for its spirit or intent?  Did the lawmakers who wrote and passed the Anti-deficiency Act, FCRA and WIFIA’s statutes ever anticipate the possibility that a complex interaction of WIFIA loan terms, tax-exempt market rates and large public agency borrowers might check all their prudential boxes yet result in an unintended and massive cost to taxpayers?  Of course not.  And nothing prevents them from revisiting the law or its interpretations with better understanding of the implications of that interaction.

The FCRA Re-Estimate Loss Ratchet – A Predictable Outcome

I’ve often written about the tendency of FCRA re-estimates to be a “one-way ratchet of losses” when loan program borrowers have excellent financing alternatives.  If a borrower has a fixed-rate loan commitment from a loan program that can be drawn over a long construction period and cancelled without penalty at any time, they’ll only draw the loan when it’s cheaper than their current debt alternatives.  WIFIA borrowers are almost exclusively large, highly rated public agencies with excellent access to the tax-exempt bond market, where yields are close to, or even below, the Treasury curve on which WIFIA loans are priced.  Hence, WIFIA loans will only be drawn when their interest rates are below the Treasury rates that are used to calculate FCRA re-estimates, resulting in a loss.  Re-estimates gains at drawdown will be rare enough, but the loss ratchet doesn’t stop there — drawn loans that are more costly than the borrower’s then-current bond alternatives will likely be quickly refinanced.  The portfolio will relentlessly accumulate only those loans that result in a deficit between interest revenue and Treasury funding cost.  Regardless of other glittering metrics, it is a portfolio of interest rate losers.

And not small-time losers, either.  The deficit will be far larger than the Congressional appropriations authorized for the portfolio precisely because the borrowers have such high credit quality.  WIFIA’s discretionary appropriations are primarily allocated to a reserve for projected credit losses, which will be infrequent and minor for highly rated public agencies.  These losses are expected to be less than one percent of the loan amount, which in effect allows a leverage of discretionary appropriations (net of administrative costs) of about 100:1.  But funding losses arise from the entire loan amount.  In effect, the high leverage of the Program’s portfolio, combined with the loss ratchet described above, will result in funding losses that are orders of magnitude higher than discretionary appropriations. 

Obviously, this outcome was not the intent of the lawmakers who passed WIFIA’s statutes.  Did WIFIA’s program managers cynically plan to exploit a FCRA loophole and create a massive portfolio of loss-making loans to highly rated public agencies because loan volume itself is a measure of bureaucratic success?  Almost certainly not.  More likely, they just went with what apparently produced apparently great results in complete compliance with the law and didn’t seem to cost any budget resources.  But original intent, after five years of actual outcomes, is not especially relevant.  The FCRA loss ratchet is now a predictable outcome if WIFIA persists in lending to public-sector borrowers with excellent financing alternatives. The important question is whether anything can or should be done about it.

Re-examining the Law, and Application Thereof

Predictable losses from a federal loan program are naturally consistent with many policy objectives.  But such subsidies require discretionary appropriations – that’s the point of precisely calculating the future expected credit losses of WIFIA loans and allocating current appropriations as a reserve to cover them.  Credit losses are of course a well-understood aspect of making loans, especially for public-sector loan programs that are intended to support worthy but risky borrowers that might not have great access to private-sector alternatives.  Likewise, funding losses from loans are that explicitly priced at less than Treasury yields for policy objectives will require discretionary appropriations.  The FCRA loss ratchet, however, is a relatively complex financial concept that will apply in specific circumstances, and (so far) apparently in scale only at the WIFIA Program.  Apart from the specific losses at WIFIA, is it a problem worth focusing on with respect to fundamental, permanent solutions?

I think it is, for two reasons.  First, if there is a Solyndra-style ideological or political attack, bureaucratic fear will probably curtail interest rate management products at the Program and chill innovation in this area (or anything remotely related) at other programs.  That’s aside from the general discrediting effect on all future infrastructure loan programs.  But if WIFIA’s specific losses are characterized in terms of a larger problem encountered in a new (and very promising) area of federal infrastructure support, and fundamental fixes are being developed, a principle-based defense will deflect a (supposedly) principle-based attack.  The approach also goes beyond WIFIA.  Developing fundamental solutions to loan program budgeting ambiguities are both a useful short-term defensive tactic and a valuable long-term framework to keep the focus on loan program potential.

Second, even if there isn’t an attack, reliance on budgeting loopholes should be discouraged at WIFIA and other infrastructure loan programs.  Resource misallocation is ultimately not sustainable, regardless of the ability to hide federal losses, because it doesn’t produce real-world improvements.  Offering a free version of a loan feature that infrastructure agencies would ordinarily buy in the private sector might be a fun and easy way to build loan volume at a program, but it’s essentially a minor subsidy payment, unlikely to change much in terms of real-world infrastructure outcomes.  If the actual cost of that “free” feature had to be considered, other ways to deploy federal lending strengths would need to be explored by a program and its stakeholders, a path more likely to have a transformative effect.  Impactful policy development at loan programs requires budget discipline.  That discipline and its policy effects are improved if budget guidance is updated with depth and nuance when new situations arise.  

Fundamental solutions should start with the most fundamental law in this case, the Anti-deficiency Act.  Here is a core passage from a 2006 GAO report, Principles of Federal Appropriations Law:

The combined effect of the Anti-deficiency Act, in conjunction with the other funding statutes discussed throughout this publication, was summarized in a 1962 decision. The summary has been quoted in numerous later Anti-deficiency Act cases and bears repeating here:

“These statutes evidence a plain intent on the part of the Congress to prohibit executive officers, unless otherwise authorized by law, from making contracts involving the Government in obligations for expenditures or liabilities beyond those contemplated and authorized for the period of availability of and within the amount of the appropriation under which they are made….”

WIFIA loan commitments that, when drawn as loans, will predictably cause funding losses far in excess of the Program’s appropriations would clearly be prohibited if such losses were not in effect authorized by FCRA law.  I can see why there is an exception to Act’s overall “plain intent” for certain types of federal laws, those pertaining to core obligations like Social Security for example.  But does FCRA naturally belong in that category?  The lawmakers who wrote and passed the FCRA law obviously didn’t think that making loans was a core obligation of the federal government to its citizens.  In fact, FCRA is all about correctly assessing the amount of appropriations required to cover the cost of federal loans.  Yet there’s clear and succinct language in FCRA that provides ‘permanent indefinite budget authority’ (PIA) for interest rate re-estimates.  In effect, FCRA’s exclusion here creates the legal authorization for the Anti-deficiency Act’s exclusion — which in turn exempts WIFIA’s FCRA re-estimate loss ratchet from budget discipline and has allowed the Program to rack up billions in future mandatory appropriations.  How did PIA even get added to the FCRA law?  What was the intent?  How was it meant to be used?

FCRA law itself doesn’t give any indication on the intent of the PIA provision.  But that law is also silent on another budget topic that has arisen at WIFIA, the extent to which loans to federally involved projects can receive FCRA treatment.  To develop criteria for such loans, OMB was instructed by Congress to refer to a foundational document of current federal budgeting, the 1967 Report of the President’s Commission on Budget Concepts.  There’s an entire chapter on federal credit programs in that report which outlines the principles on which FCRA was ultimately closely based.  That seems like the right place to start.

The Commission makes it very clear that the full funding cost of federal loans needs to be calculated and paid for (emphasis added):

It is the Commission’s recommendation that the full amount of the interest subsidy on loans compared to Treasury borrowing costs be reflected and specifically disclosed in the expenditure account of the budget, and furthermore, that it be measured on a capitalized basis at the time the loans are made.

The phrase “at the time the loans are made” is more ambiguous.  Is the loan ‘made’ at the time of commitment?  Or when the loan commitment is drawn?  I think the Commission meant the latter.  Logic would suggest that “Treasury borrowing costs” are most relevant when the loan is being funded, not at the time of commitment, though the basis of future drawdowns (i.e., the loan’s fixed interest rate) is defined at that point.

Yet a loan commitment does create a federal obligation whose full cost requires the allocation of budgeted resources when the commitment is executed.  Since loans intrinsically include future variable factors, their cost is necessarily an estimate.  The Commission put some thought into that, most likely in connection with credit losses:

The Commission also recommends that effective measures be developed to reflect (in the expenditure rather than the loan account of the budget) the further subsidy involved in the fact that Federal loans have a larger element of risk than borrowing. This should be done by creation of allowances for losses and making appropriate credits to those allowances and charges to expense as new loans are extended.

Importantly, note that although the risk of future funding losses is not explicitly mentioned here, such losses are not conceptually different than future credit losses in the context of the Commission’s recommendation.  Federal loans that are drawn many years after commitment have repayment and funding risk.  The extent to which these risks will result in future losses is obviously not known at the time of commitment.  Hence, “effective measures” must be developed to reflect the expected cost as accurately as possible.  Such measures, customized for each borrower and individual loan, have been developed and are in full operation at WIFIA for individual loan credit losses.  “Allowances” for such losses are where the Program’s discretionary appropriations go.  Could equivalent measures be developed for estimating funding losses due to the length of time over which the WIFIA loan commitment can be drawn and considering the borrower’s financing alternatives?  The short answer is ‘yes’ even in the early-computer times of 1967 and certainly now.  Well, why didn’t it happen?

The clear intent of the relevant laws was to include the full cost of a loan in the discretionary budget.  Perhaps the lawmakers didn’t think that post-commitment funding losses would be material, and that the most efficient way to deal with such ‘noise’ was to dump it in an off-budget account with PIA, segregating it from the true budget ‘signal’.  That perspective makes sense given the typical characteristics of federal loans:  

• Federal loan program borrowers don’t usually have good private-sector debt alternatives, never mind excellent ones at near-Treasury rates.  Isn’t the lack of private-sector alternatives the point of loan programs in the first place?  Loan commitments to the typical federal borrower will be drawn (and sincerely appreciated) regardless of whether Treasury rates have risen of fallen.

• The time between loan commitment and drawdown is usually relatively short.  Infrastructure loan programs are rare and relatively recent in the total federal picture.  And even for these, construction loan commitments will be drawn as quickly as possible, regardless of alternative interest rates, because the project simply needs the money.  Likewise, construction loan commitments will only be cancelled if the project failed to proceed, an outcome not usually correlated with interest rates.

• With the two above factors, interest rate re-estimate gains will be about as frequent as losses.  On a portfolio basis, these would generally balance each other over interest rate cycles.  The re-estimate off-budget account might rise and fall at various points, but overall, it will tend to zero.  In that context, FCRA’s PIA for the re-estimate account is simply a smoothing mechanism.

As it turned out over the past five years, WIFIA Program’s typical loans have almost the opposite characteristics to those listed.  I don’t think anybody could have anticipated this outcome, least of all the lawmakers who created the Program. This is reflected in the short interest rate section of WIFIA’s statute, which clearly show the intention that WIFIA loans would at least cover Treasury’s cost of funding them, and discretionary appropriations were required solely for expected credit losses.  The lawmakers would have justifiably assumed that FCRA law, voluminous specific OMB methodology, and the overall intent of the Anti-deficiency Act and its legal framework would take care of the details.

Yet, like flowing water finding unexpected paths and causing hidden damage, WIFIA’s current FYE 2022 portfolio will ultimately cost federal taxpayers far more than Congress ever intended.  What can be done to mitigate the current damage and prevent more in future?

 A Simple, Wrong Answer

The simple answer is to limit (by law or policy) WIFIA’s ability to lend to borrowers that have efficient, near-Treasury alternatives.  The FCRA re-estimate mechanisms will then work as intended.  Future re-estimate gains will occur, and higher yielding drawn loans will stay in the portfolio.  That will stop further accumulation of losses and, over time, could offset some of the FYE 2022 portfolio’s future losses.

As noted above, that limitation is a naturally occurring element of basic policy rationale for federal lending in the first place – if borrowers have such alternatives, they don’t need federal loans, right?  It’s also indirectly indicated in OMB’s Circular A-129, which (among other guidance for federal credit) requires loan program reviews to address:

Why [program objectives] cannot be achieved without Federal credit assistance, including:

(I) A description of existing and potential private sources of credit by type of institution, and the availability, terms and conditions, and cost of credit to borrowers.

(II) An explanation as to whether and why these private sources of financing must be supplemented and/or subsidized.

The simple answer, however, is wrong, or at least an unnecessarily bad policy direction.  In the US, almost all basic public infrastructure is built and operated by state and local public-sector agencies which inevitably have high credit ratings and access to the tax-exempt bond market.  To take these borrowers out of the federal credit policy equation would practically guarantee that essential infrastructure loan programs will not have a significant impact in this critical economic sector.

A Better Approach

Instead of excluding important infrastructure borrowers, improve the programs.  For that objective, WIFIA’s specific issues provide a good framework to refine budgeting methodology and deepen policy rationale for highly rated federal borrowers – a lemonade from lemons approach.  The Program’s past and current travails with federally involved project criteria also provide some guidance as to how the re-estimate issue could play out, but this time towards a positive goal.  A Congressional directive to examine WIFIA’s future funding losses explicitly in the context of solutions that will allow the Program to keep lending to highly rated public agencies would be ideal.  For the immediate future, perhaps budget oversight agencies like OMB and CBO should be involved.  There’s a pre-emptive element to this, of course, but more optimistically they might be helpful in developing solutions that they know they’ll need to sign off on anyway for political reasons.

What might near-term solutions look like?  I can think of three directions that would seem to address WIFIA’s specific issues while laying the groundwork for more universal principles for improving federal infrastructure loan programs:

• Disclosure:  As always, more disclosure is a good place to start, especially with respect to program cost.  But it should not be limited to (or even focused on) FCRA estimates of economic cost, which are important but easily misunderstood in a political context.  Instead, other types of cash-based analyses in the most relevant budgeting period can be highlighted, and other factors (e.g., WIFIA loans’ effect on tax revenue) included.

• Treasury hedging:  If federal infrastructure loan programs are going to offer interest rate management products, Treasury should explore ways to actively manage their risk and cost to taxpayers.  In some ways, WIFIA’s current funding losses could have been substantially reduced if Treasury (working with input from the Program) had hedged loan commitment exposure.  WIFIA itself might not justify an extensive operation, but in the expectation of much-increased federal infrastructure lending, a pilot-scale initiative would be valuable for larger-scale development.  Active hedging at Treasury will likely incur a measurable budgetary cost, but that’s as it should be.  Since Treasury will have vast economies of scale and efficiencies relative to any private sector hedging alternatives, true economic gains are possible.

• Develop unique loan features:  This is the most important and substantive type of solution for both the budget and policy issues of federal lending to highly rated public agencies.  If programs can offer loans with valuable features that have no near-alternatives in the private-sector (a very long term relative to the municipal bond market, for example), borrowers will be reluctant to cancel commitments or refinance drawn debt of such loans simply because interest rates have fallen.  To the extent that such features are also based on relative federal lending strengths (there are a few, and they’re real) net economic gains are also possible.  In many ways, this ought to be overall goal of improving federal infrastructure loan programs.

I’ve written several long posts on this topic. You can read this short one instead.

It’s almost certain that WIFIA’s current $16 billion portfolio will require future mandatory appropriations far in excess of discretionary appropriations received by the program so far. There are various ways to look at the issue, as the chart above shows. This requires a narrative.

The shortest bar is the place to start. It reflects the portfolio’s projected cash requirements for the next ten years, (1) using CBO’s own economic assumptions and its standard budget time horizon, and (2) including increased federal tax revenue from WIFIA loans’ displacement of tax-exempt debt, per JCT methodology.

This approach results in about $700 million of mandatory appropriations over 2023-2032. That’s only $70 million a year, or less than one-half of one percent of the portfolio annually.

See? Not a problem.

A prior post described an analysis of the estimated economic cost of WIFIA’s $16 billion portfolio of loans and loan commitments at federal FYE 2022. For various reasons outlined there, a simple estimate of the cash cost of funding the portfolio is probably a more pragmatic approach. This post starts to sketch that out.

WIFIA has built its $16 billion portfolio with 92 loans over the last five fiscal years. I assume that the average WIFIA loan has a weighted-average life (WAL) of about 20 years and therefore is executed or re-executed with an interest rate corresponding to then-current 20Y UST yield. My portfolio model starts with the reported loan amounts and execution dates. The model assigns each loan an interest rate equal to the 20Y UST on its execution date.

My portfolio model shows a simple average interest rate of 1.96% for the 92 loans. WIFIA’s website reports a 2.00% average, which I assume is also simple, given its purpose.  This seems to confirm my estimates of the portfolio’s basic characteristics.

The weighted-average interest rate (WAIR) is a more useful metric for estimating future portfolio interest revenue because it weights individual loan amounts. I estimate the WAIR at FYE 2022 to be 1.86%, a little lower than the simple average due to loan volume in the second half of 2021.

The chart below summarizes the data I’m using on a monthly basis FY 2018-2022.

The portfolio model at FYE 2022 is the basis for projecting future loan interest revenues and Treasury funding cost. Assumptions for three additional primary factors are necessary, (1) schedule of drawdown of remaining loan commitments and overall loan portfolio principal amortization, (2) the WAIR each year, and (3) Treasury’s average interest cost to fund the outstanding loan balance each year.

Simplified Primary Assumptions

In a detailed model, each loan’s drawdown, amortization schedule and interest rate would be considered, and projections would be based on the aggregate. The WIFIA program certainly has that data, and even from publicly available information (e.g., press releases about a project progress, borrower financial reports, etc.) a lot could be estimated. For this exercise, however, I’m going to make some highly simplified assumptions about the portfolio’s characteristics as a whole. This will of course limit the precision of the results, but I think they’ll still be accurate enough for the purposes discussed in more detail below.

For drawdown and amortization, I start with the assumption that the portfolio is 75% undrawn at FYE 2022. This appears to be consistent with EPA’s estimated numbers in the 2023 White House budget. The portfolio is assumed to be fully drawn in 2028, with 35-year straight-line principal amortization starting in 2029. The portfolio is fully amortized in 2064.

I assume that the portfolio’s WAIR in any year will remain at 1.86%. This certainly won’t be the case in reality, as various loans amortize in different ways, and the annual WAIR will fluctuate. But I think on average over the portfolio’s full term, it should be accurate enough.

Assumptions about Treasury’s interest cost to fund the portfolio are the main difference between economic and cash projections. Obviously, Treasury does not actually match fund its ‘investments’ in the WIFIA portfolio by issuing zero-coupon bonds, the methodology used in FCRA economic evaluations. Instead, portfolio funding is sourced from overall debt issuance, and will be reflected in an amount of additional federal debt equal to the portfolio’s outstanding at any point. Needless to say, the size of WIFIA’s portfolio is completely immaterial to the overall level of federal borrowing, so it won’t impact anything at Treasury. In that context, the cash interest cost of funding the portfolio each year will simply be a very tiny slice of the average overall cost of federal borrowing.

The average interest rate on federal debt depends on a huge number of factors and projecting it for the next forty years is a difficult mission, to say the least. Fortunately, CBO provides exactly this projection in its Long-Term Budget Outlook, at least for years to 2052. In their 2022 Extended Baseline Projection, they assume as a primary economic variable that average nominal rates on federal debt outstanding will be 2.5% for 2022-2032, 3.4% for 2033-2042 and 4.0% for 2043-2052. I use these points to create a simple linear pattern that peaks at 4.0% and declines back to 2.5% by 2064.

CBO itself makes clear that such long-term projections are in effect speculative. But as discussed further below, that doesn’t matter for the primary purpose of this exercise, which is to present WIFIA’s cost in terms of federal cash budgeting concepts. In fact, using CBO’s own numbers (whatever they are) is central to establishing a plausible case that WIFIA’s cash numbers aren’t so bad.

The chart below summarizes the simplified assumptions for our projections.

Reduced Tax Expenditures?

There is one other factor to add to the mix, especially with respect to the purpose of this exercise. If WIFIA loans create future incremental federal revenues, could those be included to offset a part of WIFIA’s interest cost?

It appears that such a ‘dynamic analysis’ is often performed by CBO but “not generally reflected in CBO’s cost estimates”. That’s probably just as well in this case because it might raise some uncomfortable questions. But I think there is one type of dynamic analysis that CBO has included in budget scoring for WIFIA legislation — the impact of WIFIA loans on the issuance of tax-exempt debt.

I’ve written about this several times over the last few years. The most succinct summary is a one-pager in the form of a letter to the Joint Committee on Taxation in 2021. For more background, here’s a long post on the topic and a detailed analysis. One thing to note is that JCT admits that its methodology is somewhat speculative. Like the CBO’s similar admission about its projections, that doesn’t really matter for our purposes — if they’re willing to use it for scoring, we’ll work with it, too. Note also that I’m using JCT’s methodology, not their assumptions, which appear to be incorrect with respect to WIFIA’s actual outcomes.

The basic idea is simply that WIFIA loans to highly rated public-sector agencies displace tax-exempt bonds that would have otherwise been issued, thereby reducing future tax expenditures. It’s a cash concept, so estimates can be included in this cash analysis. For the current case, I make some conservative assumptions: (1) 75% of outstanding WIFIA loans displace tax-exempt bonds, with the balance having no effect, (2) a federal tax rate of 25%, and (3) taxable substitute bonds yield an average of 2.75%. The annual reduction in tax expenditures (in effect, an increase in federal tax revenues) is equal to the annual tax paid on the substitute taxable portfolio’s income.

Annual WIFIA Revenue, Treasury Interest Cost

Putting all those factors together, it’s straightforward to project WIFIA revenue (without and with the tax effect) and Treasury’s cost. The chart below summarizes the results annually. As expected, Treasury’s cost is always higher than WIFIA’s revenues — even on a cash basis.

It’s also straightforward to net the revenues and costs, as shown in the next chart below. WIFIA cash losses are at their worst when the portfolio is fully drawn and CBO projects that average federal interest rates are high. They tail off as the portfolio is amortized and rates stabilize. For the next few years, the picture isn’t so bad either — that’s what we want to work with here.

The final chart summarizes the dollar amounts and percentages of the $16 billion portfolio as of FYE 2022. The undiscounted sums, and even the discounted numbers, don’t look great. I had hoped that a cash analysis, which uses medium-term federal interest costs, might be significantly better than the economic results, which reflect the full Treasury yield curve. CBO’s relatively high baseline projections of federal interest cost derailed that simple story. But there are other ways to look at things.

Estimating the economic cost of something requires that you look at the full extent of its impact, no matter how far in the future, and estimate its present value by appropriate discounting. That’s not necessarily true for cash cost projections, which can be easily sliced & diced to determine whether inflows match outflows for periods of particular interest.

CBO establishes one such a period of interest — the 10-year budget scoring protocol for new legislation, something familiar to all policy (and political) stakeholders in WIFIA’s future, or intended lack thereof. In contrast to the full cash results, the next ten years of WIFIA’s FYE 2022 portfolio don’t look so bad, especially if the correct tax effect is included. These can be further packaged on an annual basis. For example, “The portfolio net funding cost is unfortunate, but it’s only about $70 million a year or half of one percent of the portfolio. That’s not much of a problem for a popular program, is it?”

That’s where this cash analysis is going. The goal is not truth, but narrative. The simplified and preliminary analysis in this post needs a lot of refinement. But it shows promise for its fundamental purpose.

The lights are on.  The band plays.  Applications are submitted, loans are efficiently closed, and glowing press releases are issued.  The high credit quality of the portfolio continues.  Everything looks great above the discretionary appropriations waterline. 

But below that waterline, where the technical machinery of FCRA interest rate re-estimates operates, I think the picture is very different.  Red ink is flooding in.  WIFIA loan commitments bear interest rates that reflect Treasury’s full economic cost of funding them on the day of execution.  But actual funding will occur years later.  When rates rise, so too does the cost of funding the commitments, in a scale that dwarfs the program’s discretionary appropriations.  If rates fall?  The loan commitments are cancelled or reset lower and drawn are loans refinanced.  As currently operated, WIFIA will never have anywhere near sufficient funding gains to offset massive, ever-accumulating losses.  It is true, per the letter of the law, that such re-estimate losses are off budget, but taxpayers will bear them just the same.  Did Congress intend or even contemplate this outcome?  Is this consistent with the spirit of the Anti-deficiency Act?  Does WIFIA have only political friends and no competitors with political clout?  Government loan programs are said to be unsinkable, but they can run into trouble, as the DOE LPO’s Solyndra loan demonstrated.  And WIFIA itself was nearly defunded in 2020 over a budget issue far less substantive than this.

This post brings an analysis of the economic cost of WIFIA up to federal FYE of September 30, 2022.  Prior posts covered FYE 21 and some interim periods.  All the analyses are based on publicly available information.   The methodology is outlined in the first analysis and remains basically unchanged except for a few minor refinements.

Which Reality?   It Depends

One thing is important to note.  This post, as were prior ones on the topic, will be focused on the economic funding cost of WIFIA’s portfolio, which I think is accurately reflected by FCRA’s budgeting methodology.  The economic cost of anything is a somewhat theoretical concept.  In this case, it means an estimate of the resources that the federal government must allocate to fund WIFIA loan commitments for their full 35-year term.  The best real-world data to make that estimate at a particular point comes from the current Treasury rate curve, which can be used to price a series of zero-coupon bonds that exactly match the loan’s debt service schedule.  If the loan’s expected debt service payments cover these virtual zeroes when due, as was apparently contemplated in the Program’s legislation, then no further resources are necessary for its funding.  But if they don’t – the issue here – then the shortfall must be made up from other resources, which ultimately means taxes.  Discounting that stream of expected future taxes at Treasury’s own risk-free rates (only one thing is more certain than taxes, after all) results in a fair estimate of the present value of the nation’s resources that’ll need to be allocated to fund the WIFIA loan.  Think about it as an amount of value that will be taken out of the economy, set aside for the loan, and precluded from other uses.  Things could change, of course, for better or worse, but I think the estimate is the best snapshot of the loan’s real-world cost that you’ll get at any point.

WIFIA’s economic cost could perhaps be justified if the economic benefits of the loans equaled or exceeded it.  But I’m not sure that WIFIA loans have very much impact on water infrastructure itself since most (possibly all) of the financed projects would have gone ahead anyway.  Instead, the loans appear to have primarily facilitated indirect transfer payments in the form of slightly lower water rates for the lucky communities that got them.  It might be hard to make a serious case that WIFIA’s net economic benefits cover even the Program’s small discretionary appropriations.  But one thing is unquestionably true:  $16 billion of highly rated, bond-like WIFIA loans did not miraculously unleash billions of dollars in additional economic value for the US water infrastructure sector that was somehow overlooked by long-established local utilities, most with excellent access to infrastructure financing alternatives.  The most likely economic outcome is a deadweight loss, or close to it.

Estimates of economic reality are not the only way to look at the cost of WIFIA loans, however.  In fact, they may be the least relevant for practical purposes.  That’s demonstrated by WIFIA itself.  The Program blithely offers ‘free’ interest rate options because loan cost re-estimates, per FCRA, are completely off-budget and buried in obscure footnotes.  There’s a certain irony here – FCRA methodology does a good job of precisely estimating the economic cost of loan re-estimates, but FCRA law requires that those estimates are effectively hidden.  And the result is completely predictable: A $16 billion portfolio of high-quality loans and loan commitments was created with only about $200 million of on-budget (discretionary) appropriations by offering an unbeatable interest rate feature, the actual cost of which is likely to be about $2 billion in off-budget (mandatory) appropriations.  The Program looks fabulously successful, and in one sense I suppose it is, in terms of how bureaucratic Washington sees these things.  I doubt it was an intentional plan, however.  Most likely the Program just kept going with a loan feature that seemed to attract high-quality applicants and didn’t require any additional on-budget resources.  Further thought about it was unnecessary. 

Still, even if basically hidden, the scale of the economic cost of the WIFIA Program could be an issue in a political context.  If someone wanted to dig out the numbers, there’s plenty of soundbite material in a cost factor that’s likely to be ten times higher than the budgeted appropriations.  Add in language from interpretations of the Anti-Deficiency Act’s intent (a topic for future posts) and it would not be difficult to paint a very ugly picture.

WIFIA enjoys broad support, but it is not immune from opposition.  There’ll always be the possibility of ideological or partisan attack, which is what fueled the Solyndra uproar.  More alarmingly, the Program competes directly with the tax-exempt municipal bond market, whose lunch is eaten every time a WIFIA loan with a free interest rate option displaces an otherwise similar water revenue bond.  Their lobby can’t really make an overt attack on a program that provides state and local governments with another avenue to federally subsidized infrastructure financing.  That’s a bit too close to home.  But quietly raising a scary-sounding FCRA budgeting issue to a few key players is a perfect way to damage the competition while appearing to take a disinterested, public-spirited position.  I think such a tactic might have been behind the disproportionate reaction to WIFIA’s minor issue with loans to federally involved projects.  The Program was nearly defunded in 2020 due to a delay in publishing some technical criteria related to the tiny handful of applications from projects with federal involvement.  A bureaucratic delay?  In the middle of a pandemic?  On an issue that literally has zero economic impact?  Really?  Yes, really.  And it almost worked.  

In contrast to the sideshow of federally involved projects, the Program’s economic cost of interest rate re-estimates is both central to its current operations and huge, relative to what Congress expected the cost to be.  The numbers require some analysis, but a bond portfolio analyst wouldn’t find any difficulties in making a case if told where to look and what conclusions to draw.  In some ways, I’m surprised an attack hasn’t already occurred. 

A Packaging Alternative

Between WIFIA simply ignoring the Program’s economic cost and someone gleefully using its full impact for an agenda-driven purpose, there may be another approach that can re-package the cost issue in a softer, yet somewhat valid, way.  FCRA’s reality-based methodology is unique in federal cost budgeting, which largely utilizes cash accounting concepts.  Inserting a FCRA result into an artificial, cash budget world is arguably a category mistake given the actual degree of reality federal institutions operate in, akin to telling an undiplomatic truth at a garden party.  Instead, why not follow proper etiquette and estimate the projected cash cost of WIFIA’s portfolio?  That’ll look a bit better.  More importantly, a cash approach can be sliced and diced in ways that FCRA’s lump-sum of hard economic truth cannot.  What is truth anyway?

The next posts on this topic will start to pivot to a cash approach to WIFIA’s cost.  I have two goals in mind there.  The first is to raise some (but not complete) awareness of WIFIA’s funding cost to encourage product development at the Program.  Handing out free interest rate options is obviously an easy, fun, and effective way to create loan volume, but beyond that bureaucratic metric, it doesn’t accomplish much in the real world of US infrastructure.  WIFIA could do much more, and if it’s gently made clear to the Program and its stakeholders that the interest rate product is not in fact costless, they might try harder to find better ways to spend the same amount of money.

The second goal is to preempt, or at least to prepare a defense against, a political attack on WIFIA’s cost by having numbers out there that support a different narrative.  The cash approach is the lingua franca of federal budget policymakers and far easier to understand than FCRA, making it the right material for this purpose.

For the rest of this post, however, we’ll stay in the world of cold, hard reality.  Whatever edifices of spin need to be built for the greater good of more and better infrastructure loan programs, that reality remains the foundation.

Estimated WIFIA Portfolio at FYE 2022

Estimating the basic characteristics of WIFIA’s portfolio at FYE 2022 is straightforward.  The EPA website provides a list of closed deals, including amount and execution date.  There’s also plenty of public data about daily Treasury rates.  I’m assuming that WIFIA loans on average have a 20-year weighted average life (WAL), based on the typical amortization patterns for a 35-year project finance loan.  Per WIFIA law, a loan commitment gets an interest rate that basically corresponds to the UST rate for the loan’s WAL on the execution (or re-execution) date, hence I assume it’s the 20Y UST off Treasury’s SLG list or daily curve (they’re usually about the same).

Based on loan amounts and corresponding estimated interest rates, the weighted average interest rate (WAIR) of the portfolio at FYE 2022 appears to be 1.86%.  WIFIA’s website reports a portfolio average interest rate of 2.00%, but I believe this is a simple average, which would be closer their purpose of informing borrowers.  When I run a simple average on my numbers, I get 1.96%, which is very close.  That would seem to confirm my basic assumptions. 

The first chart breaks down loan commitments closed each month for the period in which WIFIA has been operating, the last five fiscal years 2018-2022.  Monthly averages of 20Y UST and 1YUST are also included to provide a sense, respectively, of (1) execution rates, and (2) the likelihood of the borrower using the WIFIA loan versus short-term financing for construction draws.  The lower short-term rates are, the more likely that the borrower will not draw the loan and use the rate lock as an interest rate option until construction completion.

The chart reflects WIFIA’s rapid startup in a period of huge movements in interest rates.  The last five years were certainly an interesting time to be creating a $16 billion portfolio of long-term, fixed-rate commitments, no?

It’s not surprising that there was a flurry of loans closed in the fall and early winter of 2020 when rates were hitting historic lows.  I’m sure the borrowers kept plenty of pressure on for executions and re-executions.  Volume has remained somewhat steady since, despite sharply rising 20Y rates.  Of course, planned projects will need financing on their own schedules and municipal bond yields will have risen as well.  But one factor, especially for loans closed over the last six months, might be the expectation that if rates fall over the next few years, the loan can be re-executed at a lower rate.  So why wait?  Once the loan is executed, the downside of even higher rates is capped while the upside of lower rates is still available.  I’d guess that nine of the latest loans, totaling about $2.2 billion, are likely candidates for re-execution if the 20Y UST gets back down to 2.50%.  Loan re-executions aren’t a statutory feature, but if future re-execution was a conscious expectation among these borrowers, WIFIA will have a hard time backtracking from precedents established in 2018-2021.  Most likely, the Program will continue resetting loan rates because there’s apparently no budgetary reason not to, and without that shield of fiscal prudence, it’s hard to counter unpleasant reactions from the Program’s eminently qualified beneficiaries.  How can a federal program stop giving out something that appears to cost the taxpayers nothing?  Yet in fact re-executions will lower the portfolio’s WAIR, turning the ‘FCRA re-estimate loss ratchet’ a few clicks further.

How much of the $16 billion of loan commitments have been drawn and are now funded loans?  WIFIA doesn’t directly provide any data on this, and there isn’t much information from other publicly available sources.  I don’t think this is the result of confidentiality or non-disclosure requirements, but simply the fact that loan drawdowns aren’t exactly great press release material.  In effect, no one cares.  But it is an important metric for estimating the economic cost of WIFIA’s FYE 2022 portfolio going forward.  Loans that have been 90% funded are not subject to further FCRA re-estimates, which means they’re not exposed to changing rates.  In a sense, potential gains or losses are realized when the loan is funded (in effect, a transfer to Treasury) and recorded in an off-budget account, something that should show up somewhere in WIFIA’s books.

Perhaps there’s more public disclosure of WIFIA’s accounts than I’ve been able to find, but for obvious reasons (primarily widespread indifference) locating this kind of thing is not easy.  There is one hint about WIFIA’s drawn loans in the White House 2023 budget published in the spring 2022.  In the EPA technical appendix, there’s a section for the WIFIA Program in which some numbers are presented in the federal budgeting format, a complex accounting language of its own.  From what I can understand, EPA estimates that the Program will have drawn loans of $3.8 billion and re-estimate losses of $140 million at FYE 2022.

The $3.8 billion, or 25% of the portfolio, seems plausible.  The amount would include earlier loans to smaller projects that are at or near completion, and whose smaller borrowers have fewer short-term financing alternatives.  Big projects with big borrowers, as well as recent loans, would make up the other 75%.  Or at least that was a plausible picture when the estimates were put together, perhaps early in 2022.  With the historically sharp rise in short-term rates this year, much may have changed, as drawing on WIFIA loans previously being kept as an option suddenly looks like the cheapest source of financing for construction draws.  It’s possible or even likely that the portfolio’s drawdown percentage was in fact much higher on September 30^th, but for now I’ll work with the 25%.

The $140 million of re-estimate losses (presumably solely from rising interest rates) looks lower than I would have expected.  However, as noted above, funded loans are likely to be with smaller, earlier borrowers who were drawing for construction costs before 20Y UST rates nosedived.  It’s also an EPA estimate which may have changed in the last six months, and there are likely other unfathomable factors at work, too.  Like the drawdown percentage, I’ll work with the number for now.  It’s worth noting that this loss is still about 4% of the funded loans, or over four times WIFIA’s typical credit subsidy of less than one percent.  Not huge in dollar terms, but definitely headed in the wrong direction.

Those assumptions leave 75% of the portfolio, or about $12 billion, exposed to changing rates.  For simplicity, I’ll assume that these loan commitments have the same WAIR, 1.86%, as the overall portfolio, though it’s probably lower.  The economic cost of funding these loans is the center of the analysis.

Before getting into the FCRA results, it’s worth doing a quick reality check.  Imagine you’re a bond portfolio manager doing a ‘mark-to-market’ (i.e., calculation of unrealized gains or losses) estimate for FYE value-at-risk reporting.  You’ve got a $12 billion portfolio of very high-quality, fixed-rate 35-year bonds with a weighted-average yield of 1.9% and a WAL of 20 years.  The 20Y UST at your FYE is 3.95%.  What’s the very best price you could expect, however theoretical, if you sold the entire portfolio that day?

The answer, which you can calculate with a two-line spreadsheet, is about $9 billion, for a $3 billion loss. The result could change in future of course – rates could fall, and your loss would be reduced.  They could keep rising, too, however, and it’ll get worse.  The $3 billion loss estimate is simply an estimate of risk at one point.  Still, you should expect losses, potentially big ones, if you’re committed to sell the portfolio over the next few years. Senior management will not be pleased.

Potential FCRA Re-Estimate Losses

FCRA methodology essentially does the same thing as pricing a bond, by discounting future payments to a present value.  Its process can be more precise because the appropriate discount rate need only to be derived from the US Treasury curve, as the federal government is the sole financing source (and hence ‘buyer’) of WIFIA loans.  The ‘purchase’ occurs when the loan commitment is funded, creating a federal investment that is amortized with loan revenues – or mandatory appropriation for taxes, if projected revenues are insufficient.  The exact loss to be covered by taxpayers is realized and recorded when the loan is funded.  As noted above, about $140 million of such losses have already been realized in funding $3.8 billion of loans.

For the remaining $12 billion of loan commitments, we can perform a FYE mark-to-market exercise very similar to the bond portfolio manager’s doleful task described above.  The portfolio characteristics are basically the same, in terms of amount, WAIR and WAL.  The 20Y UST was 3.95% on September 30, 2022, the federal FYE.  A few things are different.  FCRA methodology refines the discount rate using a zero-coupon curve derived from the overall UST curve for that day, called the ‘single effective rate’ or SER.  My estimate of the FYE SER was 3.78%, which was used for the analysis.

Just like the portfolio manager considering a hypothetical sale of the entire portfolio on the FYE date, we can ‘mark-to-market’ what the re-estimate losses would be in the hypothetical case that all $12 billion of the remaining loan commitments were drawn at FYE.  That unrealized loss is about $3 billion, or almost 20% of WIFIA’s portfolio.

Twenty percent?  For a portfolio that Congress apparently thought would cost taxpayers less than one percent?  Of course, these are unrealized losses, and as such just an indicator of the risk of potential losses, not the final damage.  Interest rates are high right now primarily because the Fed is attempting to control inflation, but that won’t last forever.  And the total portfolio of WIFIA loan commitments literally cannot be drawn now or even soon because draws must track construction progress, which will take place over at least five or six years.  If interest rates fall during that period, realized losses won’t be so bad.

It’s not difficult to model a range of outcomes depending on rates. The chart below shows potential realized re-estimate losses on the FYE 2022 portfolio at various 20Y UST rates (as converted into SERs in the model).

The simplest way to look at the results is as the potential losses that’ll be realized at the average 20Y UST rate during the multi-year period of portfolio funding.  That average probably won’t be close to today’s 4% level, nor will it end up in 2% territory.  Maybe somewhere around 3%?  This average rate results in about $2 billion of re-estimates losses, or about 13% of the portfolio.  That’s better than 20%, but still has plenty of sticker shock in it.  Depending on your natural optimism or lack thereof, maybe it’s 2.5% (a loss of about $900 million, or 5%) or 3.5% (a loss of about $2.5 billion, or 16%).

Who knows what Treasury rates will do over the next half-decade? But no matter how you look at the portfolio’s cost, the fundamental point is always the same – for any realistic projection of the rates at which WIFIA loan commitments will be funded, there will be re-estimate losses far in excess of WIFIA’s discretionary appropriations.  WIFIA’s FYE 2022 portfolio is irreparably holed beneath the economic cost waterline. The only question is whether the volume of red ink will sink the ship.